How Generative AI and Automation Are Transforming Healthcare Contact Centers


The idea of software creating and delivering content to patients and plan members would have been unthinkable in the recent past. Now it is becoming essential to a good customer experience for patients.

The healthcare consumer experience is far too fragmented for many Americans today; it's frustrating and impersonal, leaving patients feeling as if they’re being passed around a faceless bureaucracy and treated like case numbers rather than individuals.

In a recent Accenture Global Healthcare Consumer Study, 66% of respondents reported a negative experience while accessing and managing care. Poor patient experiences cost healthcare organizations in the form of lost business. More than thee-quarters (78%) of patients in the Accenture survey who switched providers cited navigation issues as a motivating factor.

Healthcare IT leaders are aware of the customer experience (CX) problems facing patients. A 2022 survey showed that 87% of healthcare IT leaders report “moderate to extreme” challenges with fragmented patient journeys. These challenges are being exacerbated as the healthcare system adopts new technologies that add even more complexity for consumers, who are becoming increasingly involved in managing their care.

For provider and payer organizations, ensuring a superior CX for patients is becoming a matter of survival against new market entrants and streamlined care delivery models at large and vertically integrated competitors. That means meeting the needs of healthcare consumers on their terms, according to Jeff Sturman, senior vice president and chief digital officer of Memorial Healthcare System in Florida.

“We have to create multichannel ways for communicating with patients,” he says. “We can’t expect consumers or patients to always call our contact center. We need to be able to text with them, we need to be able to chat, and we need to create automation where automation makes sense. Probably most importantly, we need to create more self-service capabilities.”

Slow response time and agent burnout

Understaffed healthcare contact centers are being buried under an avalanche of calls from patients and plan members. Memorial Health System, for example, receives 100,000 calls a month from patients, Sturman says. When call centers are overwhelmed, patients, caregivers, and agents can become frustrated, even angry.

When a patient or plan member’s call isn’t answered in a timely fashion, that person may hang up without having made a needed appointment or getting answers to questions about test results or coverage for specific procedures. Delayed care and poor patient access experiences can lead to worse health outcomes and higher overall healthcare costs that can, in turn, impact business outcomes for organizations that lose patients to competitors who offer a more convenient experience.

In addition, healthcare contact center agents who lack the proper tools to successfully do their jobs of helping callers quickly become burned out and quit, putting even more strain on remaining agents and further decreasing operational efficiency that can ruin the patient or member experience. In a 2022 survey by the International Customer Management Institute, 57% of responding contact center professionals reported higher turnover in their organizations from the previous year. In most healthcare contact centers, automation is needed to keep up with consumer inquiries and maximize the efficiency of available staff rather than to contain staffing costs.

Transforming the contact center experience with Generative AI

These challenges surrounding healthcare consumer experience — disconnected systems, staffing shortages, and rising patient expectations — can seem intractable. And any veteran leader in the industry will tell you that new technology is never a magic wand that can resolve them instantly. So why does the emerging use of generative artificial Intelligence (AI) in enterprises across industries hold so much promise for healthcare, and in the contact center specifically? And how is generative AI going to be any different from the AI that has been tested and used in healthcare for patient access, revenue cycle and other critical consumer experience workflows for years now?

Generative AI and the advances we’ve seen from large language models (LLMs) represent a major advance in how humans can automate processes and tasks that involve language — reading and listening to it, understanding it, conversing with it, and analyzing it. Since contact centers are hubs of important conversations between healthcare organizations and their consumers, they are a natural focus point for organizations’ generative AI efforts. And while AI has been used for years in healthcare contact centers to improve self-service, agent efficiency, and conversation analysis, generative AI enables transformative improvements in those areas.

AI and machine learning help agents and consumers by pulling together patient information from disparate sources such as texts, call logs, and electronic health records and presenting the data in an easily consumable fashion. Generative AI can look at data in a range of formats, or across several knowledgebase articles and craft answers to the precise question from the patient or member. This saves agents from having to seek out siloed data from across the network while a patient or member is waiting on the phone. Generative AI makes self-service voice and digital bots much more accurate and helpful.

Generative AI also can be used to analyze patient/member language and tone and then offer agents advice on what to say (and how to say it). This is especially helpful for newer agents who lack experience dealing with a particular topic. Furthermore, generative AI can be used to auto-summarize transcripts of patient calls and chats, work that without AI takes up a great deal of time. AI is also a powerful way of analyzing call transcripts to assess agent performance and provide actionable feedback for improvement.

Automation of simple but time-consuming tasks in the contact center, such as making and confirming appointments or resetting a portal password, can free agents to fully engage with consumers who have more complex needs. When agents interact with and focus their attention on patients or members, both consumers and agents are more satisfied with the experience, and the odds of successful outcomes increase.

“If I can put more time into actually having a personalized interaction with a consumer versus doing what I'll call administrative stuff of note taking, that's going to be of huge value,” said Sturman. “Because now I'm taking agents and freeing them up from, again, administrative burden, getting more access, answering the call more, texting more, chatting more.”

Self-service functionality in a healthcare contact center offers consumers a way to accomplish their goals without having to engage in a conversation (and possibly wait on hold) with an agent. Enabling self-service for patients/members on voice and digital channels both dramatically reduces call volume, easing the workload for contact center agents and improving the CX in the process.

Since integrating automation, multichannel options, and AI self-service capabilities into its contact center, Sturman says, Memorial Healthcare System has lowered its contact center “speed to answer” time to 43 seconds today from more than two and a half minutes a year ago.

The ability to leverage digital tools in a way that meets the expectations of healthcare consumers for a seamless experience “is not only a differentiator, but it’s critically important for the future of healthcare,” he says.

Generative AI in healthcare requires guardrails

Of course, generative AI is a new technology for enterprises, and the idea of software creating and delivering content to patients and plan members would have been unthinkable in the recent past. The key to the success and long-term viability of generative AI in the contact center, or anywhere in healthcare, will be how platforms enable organizations to build, train, monitor, and update models quickly and safely.

Human-in-the-loop systems where generative models can be trained and simulated prior to any consumer-facing deployment will be essential. And robust monitoring and observability tools are required to discover edge cases and potential problems or inefficiencies. And even AI itself should be used to constantly evaluate the performance of different automations across the contact center, recommending optimizations and corrections faster and more thoroughly than any human supervisor could. While generative AI promises impressive conversational AI and powerful agent support tools to unlock better outcomes and experiences, it will be essential for healthcare organizations to leverage capabilities to deliver them efficiently, effectively, and responsibly.

Patty Hayward is the general manager of healthcare and life sciences at Talkdesk.

Related Videos
Lawrence Eichenfield, MD, an expert on atopic dermatitis
Video 5 - "Obstacles in Adapting Diabetes Technology to Individual Needs" - 1 KOL is featured
Lawrence Eichenfield, MD, an expert on atopic dermatitis
Lawrence Eichenfield, MD, an expert on atopic dermatitis
Video 4 - "The Impact of Continuous Glucose Monitors & Digital Solutions on Diabetes Care"
Video 3 - "The Pivotal Role of Patient Engagement and Education in Achieving Optimal Diabetes Outcomes"
Lawrence Eichenfield, MD, an expert on atopic dermatitis
Lawrence Eichenfield, MD, an expert on atopic dermatitis
Video 2 - "Addressing Coexisting Conditions: Keys to Comprehensive Diabetes Care"
Related Content
© 2024 MJH Life Sciences

All rights reserved.